ORCL Thesis

SharkWater Trading

ORCL: The $638 Billion Backlog Nobody Wants to Own

Oracle has lost more than half its market value while its contracted revenue grew fivefold. One of those two facts is wrong.

August 28, 2026  •  Equities  •  Ticker: ORCL  •  Catalyst: Q1 FY27 earnings, September 4

BLUF

Oracle closed August 27 at $151.94, roughly 56% below its October high of $345.72, on a $638 billion backlog and 18.5x forward earnings. The market has stopped valuing ORCL as a software company and started pricing it as a leveraged data center developer with one very large counterparty. That reframing is fair. The magnitude is not. This is a staged contrarian long, entered small ahead of the September 4 print and added on confirmation, with a hard invalidation below $137. If you cannot tolerate a retest of $114, you do not have the position size right.

Last close$151.94

52-week range$114 - $346

Forward P/E18.5x

RPO backlog$638B

FY26 free cash flow-$23.7B

FY27 net capex$70B

What Actually Happened

Oracle spent eleven months as the purest listed expression of the AI infrastructure trade and then spent the following ten months being unwound as exactly that. The stock peaked at $345.72 last October on the back of the OpenAI relationship and a backlog that went vertical. It now trades at $151.94.

Nothing in the reported fundamentals broke. The June 10 fiscal 2026 close was a genuinely strong print: full-year revenue of $67.4 billion up 17%, cloud revenue of $34.0 billion up 39%, cloud infrastructure of $18.1 billion up 77%, non-GAAP EPS of $7.63 up 27%. Fourth quarter OCI revenue grew 93% year over year. Remaining performance obligations finished at $638 billion, up $85 billion sequentially and up 363% year over year.

What broke was the financing story. Oracle burned $23.7 billion of free cash flow in fiscal 2026, generating $32 billion of operating cash flow against $55.7 billion of capital expenditure. Fiscal 2027 capex guides to roughly $70 billion, funded by a planned $40 billion raise that includes a $20 billion at-the-market equity program. Debt cleared $100 billion. On August 5, Oracle's five-year credit default swap spread hit a record. In late June the stock had its worst week since the 2001 dot-com unwind.

That is the whole story in one sentence: the equity got repriced by the credit market, not by the income statement.

The Bull Case

The backlog is contracted, and it converts on a known schedule

RPO is not a pipeline or a letter of intent. It is signed revenue. CFO Hilary Maxson laid out the conversion curve on the Q4 call: roughly 12% inside twelve months and another 34% in the thirteen to thirty-six month window. Against a $638 billion base, that is something on the order of $77 billion recognizing over the next year and another $217 billion over the two years after. Fiscal 2027 revenue guides to $90 billion, up from $67.4 billion, which is a 34% step in a business the market is valuing at 18.5x forward earnings.

The capex is demand-driven, not speculative

The distinction matters enormously and the market is currently refusing to make it. Oracle is not building capacity and hoping tenants show up. GPU utilization ran at 97.5%. Four customers have each committed more than $8 billion. Oracle delivered over 1.2 gigawatts in fiscal 2026 and expects to deliver nearly a full gigawatt in the first quarter of fiscal 2027 alone. Management explicitly tied the $70 billion capex number to "committed customer demand reflected in our record RPO." A data center with a signed multi-year tenant at 97.5% utilization is a very different asset than an empty shell.

Some of the balance sheet risk is already moving off the books

In April, Related Digital, in partnership with Blackstone, secured $16 billion of financing to develop data centers built for OCI. Oracle is increasingly the tenant and operator rather than the owner of record. That trades some long-run economics for near-term balance sheet relief, which is exactly the trade a company in Oracle's position should be making. The bear case that treats the entire $70 billion as Oracle balance sheet risk is already out of date.

The forced selling appears to be finished

Citi's Tyler Radke reinstated a 90-day positive catalyst watch on August 26 with a $330 target, attributing the collapse to "panicked investors and mechanical selling pressure tied to credit markets" and noting that the at-the-market equity program has concluded and bond and CDS spreads are improving. Management has said it will not issue additional debt during calendar 2026. The two mechanical sources of supply, dilution and credit-driven de-risking, are both fading. Consensus across 44 analysts sits at a $246 average target.

Diversification of the counterparty is quietly underway

The Pentagon awarded Oracle an agreement worth nearly $7 billion. Neocloud demand is firming pricing. Every dollar of non-OpenAI backlog directly reduces the single largest objection to owning this stock.

The Bear Case

The unit economics on AI cloud are thin to negative

This is the most serious charge and it deserves to be taken seriously. Internal documents reported by The Information put Oracle's average gross margin on AI server rentals at roughly 16%, ranging from 10% to 20%, with one three-month stretch at 14%, meaning $125 million of gross profit on $900 million of revenue. Oracle reportedly lost nearly $100 million renting Nvidia Blackwell chips over a comparable period. Depreciation alone costs about seven points of margin. Management has guided that fiscal 2027 gross margins will step down further.

A $638 billion backlog at 15% gross margin is a fundamentally different asset than a $638 billion backlog at 60% gross margin, and Oracle's legacy database business trained everyone to assume the latter.

Depreciation schedules are an unresolved question

The entire margin structure depends on how long a GPU is assumed to earn. Stretch the useful life and margins look survivable. Shorten it, as accelerating silicon release cycles arguably demand, and the seven points of depreciation drag becomes considerably worse. Nobody outside the company knows the answer, and the answer determines whether this backlog is worth owning.

Counterparty concentration is the whole trade

The OpenAI relationship, reported at roughly $300 billion, is the majority of the story. In March, Oracle and OpenAI scrapped plans to expand a flagship data center. Reports have circulated that OpenAI may delay its IPO. OpenAI is a private company funding enormous compute commitments out of capital raises rather than operating cash flow. If that funding chain tightens, Oracle is holding purpose-built capacity and a receivable rather than a customer.

The dividend and the buyback are hostages

Free cash flow does not turn positive on the current capex plan for years. The $2.00 annual dividend, a 1.32% yield, is being paid out of a company burning cash. Buybacks are effectively suspended. Even the constructive analyst work concedes that cash flow is impaired through 2026 and that the dividend carries risk.

The credit market is not obviously wrong

Equity holders have a habit of assuming credit is panicking. Credit has a better track record than equity at identifying funding stress ahead of time. Record CDS levels on a company adding $40 billion of financing against negative free cash flow is not an irrational signal. It is a repricing of where Oracle sits in the capital structure hierarchy.

The SharkWater Take

The bear case is real and I am not going to talk you out of any of it. The margins are thin, the counterparty is concentrated, and the cash burn is genuine. But the market is currently pricing all of that plus a scenario where the $638 billion backlog substantially fails to convert. At 18.5x forward earnings on 34% guided revenue growth, you are not paying for the upside. You are being handed it in exchange for accepting the funding risk.

Here is the asymmetry I care about. If Oracle simply executes the guided plan, $90 billion of revenue and $8.05 of non-GAAP EPS in fiscal 2027, and the credit market calms down, a 25x multiple puts the stock near $200 without anything heroic happening. If the backlog converts at the pace management describes, the 2028 and 2029 numbers get silly. If OpenAI stumbles or margins compress further, $114 is on the table and lower is possible.

That is roughly 25% of downside to the 52-week low against 30% to 100% of upside, with a hard catalyst seven days out. I will take that skew, but only with a position structured to survive being early.

How I am playing it

• Starter position now, one third of target size. The stock reclaimed its 20-day ($144.65) and 50-day ($142.90) averages and cleared $149 resistance on August 27. That is enough confirmation to be involved, not enough to be full.
• Do not size up into September 4. Q1 FY27 is a binary event. Watch three things: sequential RPO direction, the actual gross margin line versus the guided step-down, and any commentary on capex pacing or financing. Guidance was 27% to 29% revenue growth for the quarter.
• Add on the reaction, not the number. A print that holds $149 on the day after earnings gets the second third. A reclaim of the 100-day, roughly $165, gets the last third.
• Invalidation is a daily close below $137. That is the identified support shelf and the level where the credit narrative reasserts control. Below there the next real reference is the 52-week low at $114.50.
• Cap it. This is a single-name bet on a financing outcome, not a core holding. Size it where a full retest of the low is annoying rather than structural.

No options structures in this post. Any spread worth putting on here has to be built off live chains, and I will not publish strikes and premiums I have not pulled from a broker.

Levels and Dates

Reference	Level	Meaning
52-week high	$345.72	October peak, the reference for the 56% drawdown
200-day average	~$172	Roughly 13% above spot, trend repair
100-day average	~$165	Roughly 10% above spot, final add trigger
Resistance	$149	Cleared August 27
Last close	$151.94	August 27, up 2.06%
20-day average	$144.65	Reclaimed
50-day average	$142.90	Reclaimed
Support	$137	Invalidation on a daily close below
52-week low	$114.50	The number you must be able to tolerate

Calendar: Q1 fiscal 2027 earnings on September 4. Investor day in late October, which is where the longer-term AI strategy and, more importantly, the depreciation and margin framework should get aired.

What Would Change My Mind

• Sequential RPO declining, or a material contract being restructured rather than added to
• AI cloud gross margins printing below the low end of the reported 10% to 20% band
• Any announcement of additional debt during calendar 2026, which management has said will not happen
• A dividend cut, which would confirm the cash flow situation is worse than guided
• Concrete evidence of OpenAI funding stress, including further data center cancellations

Fair winds and following seas.

This is analysis and opinion, not investment advice. I am not your financial advisor and I do not know your situation, time horizon, or risk tolerance. Figures are drawn from Oracle's fiscal 2026 fourth quarter results reported June 10, 2026, subsequent public reporting, and market data as of the August 27, 2026 close. Do your own work before risking capital.

IREN Thesis

SharkWater Trading  •  Equities Desk • IREN • AI Infrastructure

IREN's $4 Billion ARR Is Real. So Is the $27 Billion Bill Sitting Behind It.

August 27, 2026

Bottom Line Up Front

IREN Limited (NASDAQ: IREN) closed FY26 with $707.0m in revenue, a $702.6m net loss, and $638.8m of non-cash impairments as it scrapped its bitcoin fleet. It also closed the year with $4bn of contracted ARR against 2026 capacity and $1bn already operating as of August 26, 2026. The market cap into the print was roughly $14.5bn at $40.53.

The risk is not demand and it is not price per megawatt. The risk is that management guided FY27 capex of $25bn to $30bn against roughly $22bn of identified funding, only $14bn of which is actually committed. This is a financing trade wearing an AI costume.

What Actually Happened in FY26

IREN stopped being a bitcoin miner this year. Not gradually. Mining revenue fell from $111.2m in Q3 to $66.7m in Q4 as the company physically pulled hardware to clear floor space for GPUs. Management said on the call that mining will be effectively decommissioned by the end of December 2026. The $450.4m Q4 impairment is what that decision looks like on a balance sheet.

What replaced it is growing faster than the mining book shrank. AI Cloud Services revenue went from $16.4m in FY25 to $128.8m in FY26. In Q4 alone it hit $70.5m, or 51.4 percent of total revenue. That is the first quarter in IREN's history where the AI business was larger than the mining business.

The headline number missed. Q4 revenue of $137.2m came in under a roughly $157m consensus, and adjusted EBITDA collapsed to $19.2m from $59.5m the prior quarter. Shares closed up 2.4 percent at $40.53 and then dropped 7 to 8 percent after hours.

The FY26 Scoreboard

Metric	FY26	FY25
Total revenue	$707.0m	$501.0m
AI Cloud Services revenue	$128.8m	$16.4m
Bitcoin mining revenue	$578.2m	$484.6m
Adjusted EBITDA	$245.7m	$269.7m
Net income / (loss)	$(702.6)m	$86.9m
Non-cash impairments	$(638.8)m	n/a
Cash and equivalents (incl. restricted)	$7,619.5m	n/a
Total debt	$7,593.0m	n/a

Source: IREN FY26 results release, August 27, 2026. Balance sheet figures as of June 30, 2026, and include $1,670.3m of restricted cash tied to Microsoft GPU financing. Consensus revenue estimate is a third-party compilation, not a company figure.

The Number That Runs the Thesis

Forget the loss. The impairment is the cost of a decision already made, and it does not consume cash. The number that decides this stock is FY27 capex.

Management guided $25bn to $30bn of capital spending for FY27. Against that, CFO Anthony Lewis laid out roughly $14bn of existing cash plus committed GPU financing and customer prepayments, a further $8bn targeted in additional GPU financing or prepayments, and then, in his words, the balance coming from data center financing, operating cash flows, and corporate sources.

Do the arithmetic the company did not put on a slide. Committed funding is $14bn. Midpoint capex is $27.5bn. The unfunded portion is $13.5bn, of which $8bn is a target rather than a signature and $5.5bn is a sentence. The entire enterprise value of the company is about $14.5bn.

IREN is a boat that has pre-sold four years of catch and financed the nets against the buyers' signatures. If every buyer shows up at the pier, the math is beautiful. If one does not, the nets still have to be paid for.

Where the Funding Stands

Source	Amount	Status
Cash on hand (June 30, 2026)	$7.6bn	Held, $1.7bn restricted
GPU financing, Microsoft deployments	$3.6bn	Committed, ~6.0% rate
GPU financing, other customers	$2.8bn	Committed, ~9.0% rate
Additional GPU financing / prepayments	$8.0bn	Targeted, not committed
Data center debt, op cash flow, corporate	$3bn to $8bn	Unspecified
FY27 capex plan	$25bn to $30bn	Guided

Source: IREN FY26 results release and Q4 FY26 earnings call, August 27, 2026. The first three rows sum to the ~$14bn management describes as committed. The $2.8bn tranche includes $2.4bn led by Blue Owl and PIMCO. The fourth and fifth rows are management's stated intent, not executed facilities, and the fifth is my arithmetic on the guided range rather than a company figure. Rate figures are as disclosed by the company and are approximate.

The Part the Bears Keep Underweighting

Pricing is going the right way, hard. Three to five year contracts are now signing in excess of $20m per megawatt of IT load, with active discussions around $25m per megawatt. That is roughly a 125 percent increase since November 2025. Customer prepayments are now covering 45 to 55 percent of GPU capital expenditure, which means the customer is funding the asset before IREN turns it on.

Apply the pricing to the pipeline and you see why the equity trades where it does. The 2027 target is 0.8GW of IT load. At $25m per IT megawatt, 800 megawatts is $20bn of annualized revenue. That is not a forecast and I am not presenting it as one. It is the arithmetic of the company's own two disclosed inputs, and it explains why a $14.5bn market cap does not look absurd to the people buying it.

The Capital Structure Is Already Crowded

IREN raised roughly $19bn across the last twelve months. Inside FY26 that included $2.1bn of share issuance, $3.0bn of convertible notes, and $938.0m of financing facilities. The May 2026 convertible carries a 1.00 percent coupon, matures in 2033, and was struck at a 32.5 percent conversion premium with a capped call at $110.30 against a $55.15 reference price.

Two things follow. First, at $40.53 the stock sits 27 percent below the reference price of its own convertible, which means the capped call the company paid $201.3m for is a long way out of the money and the dilution protection is not doing much work today. Second, gross debt of $7,593.0m against FY26 adjusted EBITDA of $245.7m is 31 times. That ratio only makes sense if you believe the ARR ramp, and the market is being asked to underwrite the ramp with the debt already on the books.

There is a smaller overhang on top of it. The August 4, 2026 filing tied to the Mirantis acquisition registered up to 11,981,668 ordinary shares for potential resale, roughly 3.4 percent of the 357.4m shares outstanding.

The Bull Case

• The revenue is contracted, not projected. $4bn of ARR against 2026 capacity is signed paper with a weighted average term of about four years, spread across Microsoft, NVIDIA, Perplexity, Figure AI, Together AI, Fluidstack, and others. $1bn of it is already operating as of August 26, 2026.
• Customers are pre-funding the assets. Prepayments now cover 45 to 55 percent of GPU capex. That is the single most important line in the release. It converts a balance sheet problem into a working capital problem.
• Delivery is proven, not theoretical. Microsoft accepted Horizon 1, a 50MW liquid-cooled unit at Childress, in August 2026, and IREN achieved NVIDIA Exemplar Cloud status on GB300 NVL72. Three more Horizon phases are targeted for Q4 to reach 200MW.

The Bear Case

• $13.5bn of FY27 capex has no signature on it. $8bn is targeted GPU financing and the remainder is described as data center financing, operating cash flow, and corporate sources. Corporate sources at a $14.5bn market cap means equity, and equity at these levels is expensive.
• The operating business does not currently fund anything. Q4 adjusted EBITDA of $19.2m is down 68 percent sequentially. Mining cash flow is being switched off by December 2026 and the AI Cloud gross margin has not been disclosed. IREN is asking to be financed through the gap.
• The positioning is a loaded spring in both directions. Short interest reached 26 percent of float as of August 16, 2026, with implied volatility near 109 percent on the earnings-week expiration. A stock this crowded does not reprice gently when the financing narrative changes.

The SharkWater Take

I am long this story and short this entry. The demand side is settled. Microsoft signed for $9.7bn, the pricing per megawatt has more than doubled in nine months, and customers are wiring money before the racks are energized. You do not get that combination in a business the market has correctly valued.

But I will not own the common outright into a fiscal year where planned capex is nearly twice the entire enterprise value and a third of it is unfunded. The bear case here is not that AI demand cracks. It is that credit markets have one bad month in the middle of a $27bn build and IREN has to sell equity at whatever the tape gives it. That is a gap risk, not a drift risk, and it does not respect your stop.

The tempting trade is selling puts into 100-plus implied volatility with 26 percent of the float short. I am not taking it. When a premium is that fat on a name that has already traded from $76.87 to $22.04 inside a year, the market is not mispricing the risk. It is describing it. Defined risk or no risk on this one.

Execution Notes

Premium basis: NO LIVE CHAIN. I could not pull an executable IREN options chain at time of writing. Every options reference below is structural. Do not treat any of it as a quote, and price your own fills.

Structure: If you want exposure, a defined-risk long call spread dated past the December 2026 mining decommissioning and the $4bn ARR commissioning deadline. That window is where the thesis either proves out or does not. Short-dated anything is a coin flip on a 109 IV name.

What I am not doing: Naked short puts. Cash-secured or otherwise. The tail here is a financing event, and financing events gap through strikes.

Confirmation to wait for: Executed facilities against the $8bn targeted tranche, and Horizons 2 through 4 accepted by Microsoft to complete the 200MW. Both are checkable. Neither requires you to guess.

Invalidation: A large equity raise priced at a discount, a slip in the December 31, 2026 date for $4bn of operational ARR, or any customer prepayment being renegotiated. Any one of those turns this from a build story into a funding story, and funding stories trade at a different multiple.

Tight lines. — SharkWater

Educational and informational purposes only. Not personalized investment advice. All figures sourced as noted and accurate as of publication. Options involve substantial risk of loss. The author may hold positions in securities discussed. Do your own work.

 SharkWater Trading  •  Semiconductors • Post-Bankruptcy Equity • Earnings

WOLF: The Loss Did Not Get Four Times Worse. The Share Count Got Three Times Smaller.

August 21, 2026

Bottom Line Up Front

Wolfspeed (NYSE: WOLF) reported fiscal Q4 2026 on Wednesday, August 19, after the close. Revenue was $149.6 million, dead on the midpoint of the company's own guidance. The adjusted loss was $2.26 per share against $0.77 a year ago, and the wires called it a four times miss. In dollars, the adjusted net loss was $116.9 million versus $119.8 million a year earlier. It got smaller.

The per share number blew out because the share count fell from 155.6 million to 51.8 million in the Chapter 11 exchange. Run last year's loss through this year's count and you get $2.31 a share, worse than what just printed. The real damage sits one line down: materials revenue fell 45 percent year over year and adjusted EBITDA deteriorated to negative $62.4 million from negative $40.7 million. The stock fell for the wrong reason and landed in roughly the right place.

What Actually Printed

Everything in this table comes out of the Q4 FY26 earnings release filed as Exhibit 99.1 on August 19, 2026, for the quarter ended June 28, 2026. Prior year figures are the predecessor entity, before fresh start accounting.

Q4 FY26 (quarter ended 6/28/26)	Actual	Q4 FY25
Revenue	$149.6M	$197.0M
Power products	$106.3M	$118.6M
Materials products	$43.3M	$78.4M
GAAP gross margin	-25%	-13%
Non-GAAP gross margin	-20%	-1%
GAAP net loss	-$145.4M	-$669.3M
GAAP loss per share	-$2.81	-$4.30
Non-GAAP net loss	-$116.9M	-$119.8M
Non-GAAP loss per share	-$2.26	-$0.77
Diluted weighted shares	51,799K	155,631K
Adjusted EBITDA	-$62.4M	-$40.7M
Operating cash flow	-$54.1M	-$242.5M
Free cash flow	-$60.9M	-$454.0M
Cash and short-term investments	$1,088.6M	$955.4M

Source: Wolfspeed Q4 FY26 earnings release, Exhibit 99.1, August 19, 2026. Non-GAAP reconciliation tables, pages 10 through 12. Predecessor and successor periods are not directly comparable under fresh start accounting.

Guidance for Q1 FY27 is revenue of $140 million to $160 million, non-GAAP operating expenses of $62 million to $66 million, and non-GAAP gross margin that remains negative. No date attached to when it stops being negative.

The Four Times Miss Is A Denominator

Wolfspeed filed a prepackaged Chapter 11 on June 30, 2025 and emerged on September 29, 2025. The old equity was largely exchanged. Diluted weighted shares went from 155.6 million to 51.8 million, a factor of three.

Now do the arithmetic the wire services did not. Last year's adjusted net loss of $119.8 million, divided by this year's 51.8 million share count, is $2.31 per share. This quarter printed $2.26. On a constant denominator, the adjusted loss per share was slightly better than the year ago quarter.

The boat landed about the same catch it landed last August. What changed is that the crew went from a hundred and fifty-six hands to fifty-two. Everybody's share got three times bigger, and what they are splitting here is a loss.

None of this makes the quarter good. It makes the headline meaningless. And the headline is what moved 4.3 million shares.

There Is No Consensus To Miss

Here is the part that should stop you cold. Depending on which vendor you pulled the number from, Wolfspeed missed by four hundred percent, missed by fifty percent, or beat.

Source	Q4 EPS estimate	Q4 revenue estimate	Implied verdict
Investing.com / Motley Fool	-$0.52	$223.55M	Miss by 4x, revenue -33%
TipRanks (via CNN), one estimate	-$1.47	n/a	Miss
TradingView / FactSet feed	-$2.45	$150.00M	Beat by 7.8%
Company guidance midpoint	n/a	~$150M	In line

Source: secondary financial outlets and quote vendors, retrieved August 20 to 21, 2026. These figures are irreconcilable and I am not going to pretend otherwise. TipRanks explicitly labels its number "one estimate."

A $223.55 million revenue estimate against a company that guided to $140 million to $160 million and printed $150.2 million the prior quarter is not an estimate. It is a stale row in a database. The most plausible explanation is a model that never got rebuilt after the restructuring split fiscal 2026 into a successor period and a predecessor period.

Analyst coverage collapsed alongside the equity. TradingView's forecast page shows a target range of $25 to $30. One vendor shows a single covering analyst. Another still carried a $9.22 average target from pre-emergence work. TD Cowen weighed in after the print with a Hold and a $25 target, which is at least fresh.

Rule of the boat: a company that has been through fresh start accounting inside the trailing twelve months has no usable consensus, no usable trailing multiple, and no usable chart. Go to the filing. Everything else is a vendor guessing at which period to add to which.

What The Quote Page Will Tell You, And Why It Is Wrong

Pull WOLF up on the majors right now and you will find a trailing P/E of 1.33 next to a trailing EPS of positive $23.57. You will also find a trailing P/E of negative 2.39 on a different site. Both are artifacts of the same thing.

Fiscal 2026 splits into a successor period that lost $415.8 million and a ninety-one day predecessor stub that booked net income of $420.2 million. Sum them and you get positive $4.4 million of "full year net income," which several aggregators now display. That $420.2 million is not earnings. It is a $563.4 million reorganization gain from discharging debt in bankruptcy court. Nobody made $4.4 million.

The 52 week range is equally useless. Vendors variously show a low of $1.16, $8.05, or $0.39, because the low was set by a different security under the same ticker. Reported beta ranges from 3.9 to 5.4 for the same reason. Trailing anything on this name is measuring a tide gauge that got moved.

The Line That Should Have Moved The Stock

Strip out the denominator noise and there is still a genuinely bad quarter underneath, and it is not the one anybody wrote about.

Materials revenue fell from $78.4 million to $43.3 million. That is down 45 percent year over year. Power products fell 10 percent. So the collapse is concentrated almost entirely in the wafer and substrate business, which is the part of Wolfspeed that was supposed to be the moat. Wolfspeed does not have a scale advantage in devices. It has one in silicon carbide crystal growth. That is the segment that just halved.

Second, margins moved the wrong way in a period when management is selling a margin recovery story. Non-GAAP gross margin went from negative 1 percent to negative 20 percent. Adjusted EBITDA went from negative $40.7 million to negative $62.4 million, a 53 percent deterioration, in a quarter where revenue only fell 24 percent. Fixed cost absorption on two large fabs running well below capacity is doing exactly what fixed cost absorption does.

Third, and this is the one I would want explained on the next call: gross capital expenditure was $5 million in the quarter, down from $38 million the prior quarter and $210 million in the year ago quarter. Free cash flow of negative $60.9 million looks contained precisely because the company stopped building. You cannot ramp 200 millimeter capacity and spend five million dollars a quarter at the same time. One of those two facts has to give.

The Balance Sheet Is The Thing That Actually Got Fixed

As of June 28, 2026	Amount
Cash, equivalents and short-term investments	$1,088.6M
Long-term debt	$931.0M
Convertible notes, net	$756.5M
Total debt	$1,687.5M
Net debt	~$599M
Total stockholders' equity	$930.2M
Q4 free cash flow burn	-$60.9M

Source: Q4 FY26 condensed consolidated balance sheet, unaudited, and free cash flow reconciliation. Net debt is my calculation from the filed line items, not a company-reported metric.

A year ago this balance sheet carried $6.5 billion of current maturities and negative $447 million of equity. Today it carries $1.7 billion of debt against $1.1 billion of cash and $930 million of positive book equity, with management stating on the call that there are no maturities before 2030. Holders of $46 million of convertible notes voluntarily converted during the quarter.

At the fourth quarter burn rate that is roughly seventeen quarters of runway. Treat that number as a ceiling, not a plan: it assumes capex stays at five million dollars a quarter, which it will not if the business is actually going to grow.

What You Are Paying

At roughly $26 a share against 52.0 million shares outstanding, the market cap is about $1.37 billion. Add the $599 million of net debt and enterprise value is close to $1.97 billion. Against fiscal 2026 revenue of $665.1 million, that is roughly three times sales. Against the annualized fourth quarter run rate of about $598 million, closer to 3.3 times.

Three times sales for a business with negative gross margins is not obviously cheap and not obviously expensive. It is a bet on what the fabs are worth to somebody, and the honest answer is that nobody has marked them recently.

The Tape

Date	Price	Note
May 22, 2026	$80.82	52 week high, post-emergence
Aug 17, 2026	~$31.46	Down 10% on the session, no company news
Aug 19, 2026 close	~$29.10	Down 7.5% into the print
Aug 19, 4:10pm ET	~$29.80	Initially up 2% on the release
Aug 20 intraday low	~$24.70	Down about 15% mid-morning
Aug 20 close	~$26.00 to $26.40	Sources disagree, see below

Source: multiple secondary quote vendors, retrieved August 20 to 21, 2026. All prices approximate and unsettled. Thursday's official closing print does not reconcile across sources: reported one day declines range from 9.4% to 15.8%. Do not size anything off this table.

Two details in there are worth more than the rest of the table. First, the stock traded up two percent in the minutes after the release, then reversed hard once the numbers ran through the estimate engines. The selling was not a reaction to the filing. It was a reaction to a comparison.

Second, the options market had priced a 13.7 percent post-earnings move going in. The realized one day move came in under ten percent by most readings. Whoever sold that event premium got paid, which is the usual outcome and worth remembering the next time a binary looks expensive.

Bull Case

• The dollar loss did not deteriorate. Adjusted net loss of $116.9 million against $119.8 million a year ago. The entire per share blowout is the exchange ratio.
• Solvency is no longer the question. $1.09 billion of cash, $599 million of net debt, positive book equity, and no maturities before 2030 per management. A year ago this company had $6.5 billion of debt coming current.
• The data center line is compounding. AI data center revenue more than doubled across fiscal 2026 and grew about 20 percent sequentially in Q4, with the Gen 5 SiC MOSFET launched, a LITEON collaboration announced, and a memorandum of understanding with GE Aerospace. Small base, right direction.
• Management forecast its own business correctly. Revenue landed at the midpoint of guidance. For a company twelve months out of bankruptcy court with a new CEO and CFO, forecasting accuracy is a real signal.
• You are paying roughly three times sales for the largest western silicon carbide substrate producer, with two 200 millimeter fabs already built, capitalized, and described by management as production ready.

Bear Case

• The moat business halved. Materials revenue down 45 percent year over year. Devices are a competitive scrum. Crystal growth was supposed to be the defensible piece, and it is the piece that fell apart.
• Margins moved the wrong way. Non-GAAP gross margin from negative 1 percent to negative 20 percent, adjusted EBITDA 53 percent worse year over year. Guidance says gross margin stays negative next quarter, with no date attached to when it does not.
• The cash flow is flattered by a capex stop. Five million dollars of gross capital expenditure in a quarter, down from $210 million a year ago. That is not discipline at a company whose entire thesis is a 200 millimeter ramp. It is a pause.
• Dilution is structural. $756.5 million of convertible notes net, sitting above a 52 million share count with a float of roughly 34 million. $46 million already converted voluntarily in Q4. Every good outcome for the equity accelerates the conversion.
• Automotive visibility is poor by management's own account. The CEO flagged that automotive and industrial demand remains difficult to predict, including customer product mix changes. That is the volume that fills the fabs and fixes the absorption problem.

The SharkWater Take

I am not buying this and I am not shorting it, and I want to be precise about why, because the reasoning is not "it is risky."

The market sold Wolfspeed on a comparison to a number that three different vendors could not agree on, one of which scored the same print as a beat. That is a bad reason to sell a stock. But it arrived at a defensible price anyway, because underneath the noise the materials segment lost 45 percent of its revenue and the company is burning $62 million of adjusted EBITDA a quarter while spending five million dollars on capital equipment. Being right for the wrong reason is still being right. I am not going to fade a selloff just because the crowd got there sloppily.

What I would need to see is not complicated and it is not a lease or a contract or a partnership announcement. It is two consecutive quarters where materials revenue stops falling and gross capex goes back above thirty million dollars without the cash balance cratering. That combination would mean demand returned and management believed it enough to spend. Until then this is a well financed company with a shrinking moat business, and the balance sheet buys it time rather than momentum.

The tradeable observation here is not directional. It is that this ticker is a data quality trap. Anyone running a screen on trailing P/E, 52 week range, EPS surprise, or beta is screening a security that did not exist eleven months ago. If you find yourself looking at a positive 1.33 P/E on a company with negative gross margins, the screen is broken, not the market. That is worth more than a view on silicon carbide.

One more thing, on the vol. The event premium was 13.7 percent implied against a realized move under ten. Post-event implied volatility on a name like this usually stays elevated for a week or two after the catalyst has already passed, which is the single most reliably mispriced thing in the whole setup. I have not pulled the chain, so I am not putting strikes in print.

What I Could Not Reconcile

Thursday's closing price. Reported one day declines for August 20 span 9.4 percent to 15.8 percent across vendors, implying closes anywhere from $24.50 to $26.40. Every price in this post is approximate.

Consensus estimates. Four different figures, no way to adjudicate. I have shown all of them rather than picking the one that supports a narrative.

Debt maturity schedule. Management's "no maturities before 2030" comes from earnings call coverage, which is secondary. Bond listings for the outstanding paper show 2031 maturities on three issues. Consistent, but I have not read the indentures.

Gross versus net capex. The filed cash flow statement shows property and equipment spending net of incentive reimbursements. The $5 million gross figure comes from the earnings deck as reported by a secondary outlet, not from the release itself.

No options chain. I did not pull live quotes. No strikes, no premiums, no ladder in this post. The implied move figure of 13.7 percent is pre-earnings reporting, not a live read.

Triggers

The 10-K. Due within weeks of the release. That is where the segment detail, customer concentration, the full convertible note terms, and the going concern language actually live. The press release is a summary of a summary.

Materials revenue, Q1 FY27. Reports October 28, 2026. If $43.3 million goes to $35 million, the moat argument is finished and the three times sales multiple has to come down.

Gross capex. A return above thirty million dollars a quarter says management sees demand. A second quarter at five million says they do not.

Convertible note conversions. Each 8-K disclosing voluntary conversion tells you the notes are in the money and the share count is climbing. Track the count, not the price.

What invalidates the cautious read: a quarter with positive non-GAAP gross margin and stabilizing materials revenue. At that point the balance sheet stops being a life raft and starts being dry powder, and the multiple is arguably too low.

Fair winds and following seas.

---

Disclaimer: This post is for informational and educational purposes only and does not constitute investment advice, a recommendation, or an offer to buy or sell any security. Financial figures are drawn from Wolfspeed's fiscal fourth quarter 2026 earnings release dated August 19, 2026 and are unaudited and subject to revision in the Form 10-K. Predecessor and successor period figures are not comparable under fresh start accounting. Share prices, market capitalization, enterprise value, net debt, and per-share restatements are approximate, are the author's calculations from filed line items rather than reported company metrics, and were retrieved from secondary vendors between August 20 and August 21, 2026. Analyst estimates cited are irreconcilable across sources and are reproduced as found. The implied move figure is pre-earnings secondary reporting, not a live options quote, and no executable option prices appear in this post. Forward-looking guidance regarding revenue, operating expenses, gross margin, and debt maturities is management's, not the author's. Wolfspeed is a recently reorganized issuer with negative gross margins, negative operating cash flow, and substantial convertible obligations, and its shares are highly volatile. Verify all figures against primary filings before acting. Past performance does not guarantee future results. Consult a licensed financial professional regarding your specific situation.

 SHARKWATER TRADING  •  CLOSED-END FUNDS • INCOME • DISCOUNT MECHANICS

USA: Eleven Years, $7.01 Paid Out, And The Exact Same $6.84 NAV

August 19, 2026

BOTTOM LINE UP FRONT

Liberty All-Star Equity Fund (NYSE: USA) pays approximately 10 percent of net asset value a year, in four quarterly installments of 2.5 percent. Unlike most double digit payers, the money is real: every 2025 distribution was characterized as ordinary dividends and long term capital gains, with zero return of capital in the final tax accounting and $163.4 million designated as long term capital gain dividends.

Here is the number that actually describes the vehicle. NAV per share was $6.84 at the end of 2014 and $6.84 at the end of 2025. Over that stretch the fund distributed roughly $7.01 per share. This thing is not built to compound your share price. It is built to convert a diversified large cap portfolio into quarterly cash and hand the base back to you unchanged in nominal terms.

The only live variable is the discount, and it has moved against holders all year. Shares closed July at $5.81 against a $6.60 NAV, a 12.0 percent discount, the widest reading in at least two years. NAV is up 4.96 percent year to date. The share price is up 0.63 percent. That gap is the whole trade, and the yield has nothing to do with it.

What You Are Actually Buying

USA is a closed end fund that launched in 1986 and, per the fund's own materials, was the first closed end fund to bring multi-management to individual investors. Assets are split roughly equally among five independent institutional managers: three value shops and two growth shops. ALPS Advisors sits above them and has the authority to hire and fire.

The Board used that authority this year. Effective June 15, 2026, Loomis, Sayles & Company, with Aziz Hamzaogullari's team, replaced Sustainable Growth Advisers as one of the five managers. The current lineup is Aristotle, Fiduciary Management and Pzena on value, Loomis Sayles and TCW on growth.

The portfolio itself is the least exotic thing on my watchlist. As of the December 31, 2025 audited annual report: 137 holdings, top ten at 26 percent of the fund against 41 percent for the S&P 500, a price to earnings ratio of 26 times versus 29 times for the index, and a weighted average market cap of $795 billion versus $1.435 trillion. It is a real diversified portfolio in a market that has not paid for diversification in four years.

Where It Stands Right Now

Metric	Reading
NAV per share	$6.60
Market price	$5.81
Premium / (Discount)	-12.0%
Net assets	$2,033.2M
2026 YTD return, NAV	+4.96%
2026 YTD return, market price	+0.63%
Top 20 holdings	39.1% of equity portfolio
Largest position	NVIDIA, 5.7%
Most recent declared distribution	$0.17, ex July 16, payable Aug 31, 2026

Source: Liberty All-Star Equity Fund July 2026 Monthly Update, issuer release dated August 14, 2026, all data as of July 31, 2026. Distribution per the issuer release dated July 6, 2026. No live quote is used anywhere in this post.

The History: Four Different Payout Rates In Four Decades

The distribution policy has been in place since 1988 and the marketing describes it as a rate "that approximates historical equity market returns." Read that carefully, because it is an admission. The payout is not funded by income. It is a decision to distribute the expected long run return of the portfolio whether or not the portfolio delivered it that year.

The rate itself has moved four times, and the moves tell you more than the current number does.

Period	Annual rate	What triggered the change
1988 to Q1 2009	10% of NAV	Original policy
Q2 2009 to Q1 2015	6%	Cut in the wreckage of the financial crisis
Q2 2015 to Q3 2017	8%	Raised to realign with historical equity returns
Q4 2017 to present	10%	Board cited narrowing the discount as a goal

Source: Table of Distributions, Tax Credits and Rights Offerings, Form N-CSR for the year ended December 31, 2025, footnotes 3, 4 and 5. Q4 2017 rationale from the Fund's third quarter 2017 shareholder report.

Note what happened in 2009. The Board cut the rate by 40 percent at exactly the moment shareholders wanted the cash most. That is the correct decision for a fund that does not want to liquidate itself into a crash, and it is also the precise reason nobody should treat the 10 percent as a contract. It is a policy, and policies get changed by the same eleven people who wrote them.

The dollars follow the same story:

Year	Distributions per share	Context
2000	$1.42	Peak year of the original 10% era
2008	$0.65	Last full year before the cut
2009	$0.31	Rate cut to 6% effective Q2
2017	$0.56	Rate restored to 10% effective Q4
2021	$0.81	Recent peak
2023	$0.61	Post-2022 drawdown low
2024	$0.71
2025	$0.67
Total since 1987	$31.67	More than $4.0 billion paid out

Source: Form N-CSR for the year ended December 31, 2025, Table of Distributions, Tax Credits and Rights Offerings.

A shareholder collecting $1.42 a share in 2000 collected $0.67 in 2025. The advertised rate never changed by a basis point across most of that span. The rate is a percentage of a number that moves, so it can look permanent while the cash quietly halves. Same mechanic I wrote about in the THTA post, different plumbing.

The Eleven Year Round Trip

Here is the cleanest way to understand what this fund does.

NAV per share entering 2015 was $6.84. NAV per share on December 31, 2025 was $6.84. Over the ten years ending December 2025 the fund distributed $6.50 per share, and 2015 added another $0.51. Call it $7.01 of cash out the door against a base that finished exactly where it started.

That is the machine working as designed, and it deserves to be said plainly before the criticism starts. The fund did not eat its principal. It earned enough to pay out more than 100 percent of its starting NAV over eleven years and still hand back the same nominal base. Plenty of double digit payers cannot say that. THTA gave back $4.33 of a $20 NAV over less than three years to fund a comparable headline rate.

But a flat nominal base over eleven years is a shrinking real one. Every dollar of that $6.84 buys measurably less in 2026 than it did in 2014. For a shareholder taking the distribution in cash, there is no per share compounding at all, by construction. The entire return arrives as a check and then leaves. That is a feature if you need income and a serious problem if you were expecting a growth fund with a nice dividend attached.

The fund's own long run illustration makes the point better than I can. A hypothetical $10,000 buying shares at the December 31, 1987 closing price of $6.00, with all distributions taken in cash, grew to $61,433 by December 31, 2021. The same $10,000 with everything reinvested grew to $410,893. Roughly 5.5 percent annualized against roughly 11.6 percent, over the same thirty four years, in the same fund.

Source: Form N-CSR for the year ended December 31, 2021, Growth of a Hypothetical $10,000 Investment. Annualized rates are my calculation from those endpoints, not fund-reported figures. The reinvestment scenario also assumes the shareholder exercised all primary rights in the Fund's rights offerings, which required additional capital beyond the original $10,000.

Six percentage points a year of difference, driven entirely by whether the cash goes back in. If you own this fund and spend the distribution, you are earning something close to a bond-like number from an all equity portfolio. Know which of those two lines you are on before you buy.

Distribution Quality: The Part That Actually Checks Out

My default assumption with any double digit payer is that a large slice of the money is your own capital coming back with a bow on it. That is not what the filings say here.

The 2025 annual report states that all 2025 distributions consist of ordinary dividends and long term capital gains, and the fund designated $163,414,020 as long term capital gain dividends under Section 852(b)(3) for the year. Zero return of capital in the final tax characterization. On $0.67 per share across roughly 300 million shares, the realized gains genuinely covered the payout.

That is what a 10 percent distribution looks like when the underlying portfolio actually produced 10 percent. It is also why USA is a fundamentally different animal from a fund that manufactures its yield out of an options overlay and a NAV drip.

One flag for 2026. The fund's monthly updates through July 2026 carry standing language that, based on current estimates, a portion of this year's distributions consists of return of capital. Those are Section 19(a) estimates, not final tax character, and 2025's estimates also carried that boilerplate before the year closed clean. Do not read it as a scandal, and do not read it as nothing either. The final answer lands on the 2026 Form 1099-DIV, and the number to watch is whether realized gains cover the payout in a year where NAV total return is running under 5 percent.

Performance: The Benchmark You Pick Does All The Work

Annualized, periods ended 12/31/2025	3 Years	5 Years	10 Years
USA, shares valued at NAV	16.10%	9.16%	11.84%
USA, shares valued at market price	14.15%	8.73%	12.50%
S&P 500 Index	23.01%	14.42%	14.82%
Lipper Large-Cap Core Average	20.76%	12.71%	13.46%
S&P 500 Equal Weight Index	12.76%	10.48%	11.71%
Distributions paid per share	$1.99	$3.49	$6.50

Source: Form N-CSR for the year ended December 31, 2025, Long-Term Performance Summary. Fund returns are net of expenses and assume reinvestment at actual reinvestment prices and exercise of all primary rights.

Against the cap weighted S&P 500 this fund has been beaten badly and consistently. Three years, five years, ten years, no argument. Against the equal weight version of the same index it beat on three years, lost on five, and edged it on ten. That is the honest comparison for a portfolio whose top ten is 26 percent of assets against an index whose top ten is 41 percent.

Management makes exactly this argument in the annual report, and it is a fair one rather than an excuse. It is also cold comfort. Shareholders do not get to spend a relative return against the benchmark they wish they owned.

The line I keep returning to is the five year: 9.16 percent annualized at NAV, against a policy that distributes 10 percent of NAV every year. A fund that pays out more than it earns is shrinking its own per share base. Reinvesting shareholders partially offset that, because they buy new shares at a discount. Cash-taking shareholders do not. In a decade of average or below average equity returns, that math grinds in one direction only.

It works like a well with a pump set to draw a fixed amount every quarter regardless of rainfall. In wet years the level holds and nobody thinks about it. In dry years the level drops, the pump keeps its setting, and the change only becomes visible when somebody finally lowers a rope and measures.

The Discount Is The Only Variable You Can Actually Trade

The portfolio is a diversified large cap book. You are not going to out-analyze five institutional managers on Capital One and Broadcom. The one thing in this security that is genuinely mispriceable is the spread between what the assets are worth and what the shares cost.

Period	Premium / (Discount) range	Endpoint
Calendar 2024	+0.8% to -5.8%
Calendar 2025	+1.2% to -10.4%	-8.2% at year end
July 31, 2026		-12.0%

Source: Form N-CSR for the year ended December 31, 2025, President's Letter, and the July 2026 Monthly Update dated August 14, 2026.

Shares traded at a premium to NAV as recently as calendar 2025. They now sit 12 percent below it. That move is the entire reason NAV is up 4.96 percent this year while the shares are up 0.63 percent. Roughly four points of shareholder return were consumed by sentiment, not by anything in the portfolio.

At a 12.0 percent discount, the arithmetic of the payout changes too. Four quarters at the most recently declared $0.17 is $0.68 a share. Against the $6.60 NAV that is 10.3 percent, right on policy. Against the $5.81 share price it is 11.7 percent. You are buying the payout stream at a discount to the assets generating it.

The $0.68 figure is a run rate, calculated by annualizing the single most recently declared quarterly payment. It is not a forward yield, not a guaranteed figure, and not a fund-published number. The actual next payment will be 2.5 percent of NAV on the Friday before the next declaration.

The Reinvestment Mechanic Nobody Reads

Every distribution release contains the same sentence, and it matters more at a 12 percent discount than it did at a premium. Shares under the reinvestment plan are issued at the lower of NAV or market price on a set date, but not less than 95 percent of market value. For the August 31 payment, that reference date is August 14, 2026.

Work through what that means when the shares trade below NAV. The fund issues new shares at roughly the market price while each share represents a larger amount of NAV. Reinvesting shareholders get more than a dollar of assets for every dollar they forgo, which is a genuine benefit to them. The offset is that issuing shares below NAV dilutes NAV per share for everybody, including the shareholder who elected cash.

Share count grew from 294,774,236 at the June 2025 record date to 301,551,615 at year end 2025, roughly 2.3 percent in six months, with no rights offering since 2021. The wider the discount, the more that machinery works against the cash-taker and for the reinvestor. If you own this fund at a double digit discount and you are taking the cash, you are on the wrong side of your own dividend reinvestment plan.

Reinvestment plan terms per the issuer distribution release dated July 6, 2026. Share counts per the Form N-CSR for the year ended December 31, 2025. The dilution conclusion is my derivation from those stated terms, not a fund-reported metric.

Bull Case

• The distribution is earned, not manufactured. Zero return of capital in the final 2025 tax character, with $163.4 million designated as long term capital gain dividends. That is rare at this payout level.
• The discount is at the wide end of its own recent history. Twelve percent below NAV, in a security that traded at a premium during 2025. You are paying $0.88 for a dollar of NVIDIA, Alphabet, Wells Fargo and Visa.
• The effective payout on market price is 11.7 percent. The discount does not just offer capital upside, it mechanically lifts the cash return on every dollar you commit.
• The portfolio is cheaper and broader than the index. 26 times earnings against 29 times, top ten at 26 percent against 41 percent, 137 names. If breadth improves, relative performance improves without management doing anything clever.
• The advisor fires managers. Loomis Sayles replaced Sustainable Growth Advisers in June 2026 after a stretch of weak relative returns. The oversight layer is not decorative.
• Forty years of operating history and $4.0 billion distributed. The structure has survived 2000, 2008, 2020 and 2022. Very little on my watchlist can say that.

Bear Case

• The payout exceeds the five year return. Ten percent distributed against 9.16 percent annualized at NAV over five years. The fund needs an above average decade just to stand still per share.
• It trails its own primary benchmark. NAV returned 8.80 percent in 2025 against 15.31 percent for the Lipper Large-Cap Core Average. Three, five and ten year figures all trail Lipper and the S&P 500.
• The reinvestment plan dilutes NAV per share at a discount. Shares issued near market price while NAV sits 12 percent higher. Cash-takers absorb that.
• The rate is a policy, not a promise.The Board cut it from 10 percent to 6 percent in 2009 and the annual dollars fell from $0.65 to $0.31. That precedent exists and the same mechanism is still in place.
• The discount is widening, not narrowing. The 2017 rate increase was partly justified as a way to compress the discount. Nine years later the discount is wider than it was then, which is a reasonably direct verdict on that theory.
• Diversification is the position, and it has been the wrong one. If AI leadership stays narrow, a 137 stock portfolio with a $795 billion weighted average market cap keeps losing to the index by construction.

The SharkWater Take

I like USA more than I like most things paying double digits, and I would buy it for a reason that has nothing to do with the yield.

The distribution here is honest. I went looking for the return of capital story and it is not there, at least not in 2025. That closes off my usual line of attack and forces a different question: what do you actually own? You own a diversified large cap portfolio with a rule attached that converts its entire long run return into quarterly cash. Per share, in nominal terms, it has gone nowhere in eleven years and it is not supposed to. The compounding lives entirely in the reinvestment decision, which the fund's own 1987 illustration prices at roughly six percentage points a year.

So the yield is not an edge. Ten percent of NAV distributed from a portfolio earning 9.16 percent over five years is not income, it is a schedule. The edge, if there is one today, is that the market has marked this basket down 12 percent below the value of its contents, in a fund that was trading at a premium last year. That is a spread I can measure, and it is the only thing in this security that is plausibly mispriced.

My read: this is a buy on the discount, held for the discount, with the payout treated as carry while you wait. If the spread closes back toward the 5 to 8 percent range that prevailed through 2024 and early 2025, you collect a mid single digit capital gain on top of a market return, in a name where the downside is a diversified equity portfolio rather than a story. If the spread never closes, you have bought a slightly cheap index proxy that mails you 11.7 percent a year and compounds only if you send it back. That is a fine outcome and a poor one at the same time, depending entirely on what you thought you were buying.

What I would not do is own this instead of bonds, or size it as though the 10 percent were a coupon. It is equity. The NAV fell more than 20 percent in 2022 and the distribution fell with it, because that is exactly what the formula does.

Two things I am watching. First, the fourth quarter Section 19(a) notices and the eventual 2026 Form 1099-DIV, to see whether the return of capital estimates in this year's monthly updates survive into the final tax character. A year with sub-5 percent NAV total return and a 10 percent payout is precisely when that changes. Second, whether the discount stabilizes. It has gone from a premium to negative 12 percent in about eighteen months, and I want to see it stop widening before I believe the mean reversion story.

Execution Notes

• Anchor on the discount, not the price. The relevant number is the spread to NAV, which the fund publishes monthly and daily on its own site. A share price that looks cheap at a 3 percent discount is expensive relative to one that looks dear at 12 percent.
• You can calculate the next distribution before it is announced.The payment is 2.5 percent of NAV at the close on the Friday prior to the declaration date. The 2026 declarations landed January 12, April 6 and July 6, which points to an early October window for the fourth quarter. That cadence is my inference from this year's pattern, not a date the fund has confirmed.
• Make the reinvestment election deliberately. At a wide discount the plan issues shares at roughly market price against a higher NAV. Taking cash while the discount is double digit means paying for someone else's accretion.
• Tax location matters more than usual. The 2025 distributions were ordinary dividends and long term capital gains, so the character is reasonably friendly, but a 10 percent annual distribution generates a taxable event every quarter whether or not you wanted one.
• What invalidates the thesis: the discount widening through the negative 15 percent area without a market-wide closed end fund selloff to explain it, a 2026 tax character that comes back materially return of capital, or a Board move to cut the rate, which would be the 2009 playbook running again.

What I Could Not Verify

Three things, stated plainly rather than papered over. I did not pull the fund's total expense ratio from the December 31, 2025 financial highlights, so no fee figure appears anywhere above. I did not use a live quote: the most recent issuer-published price and NAV in this post are as of July 31, 2026, and the shares have traded for two and a half weeks since. And the Section 19(a) source-of-distribution detail for individual 2026 payments is referenced only through the standing language in the monthly updates, not through the individual notices themselves.

Fair winds and following seas.

---

Disclaimer: This post is for informational and educational purposes only and does not constitute investment advice, a recommendation, or an offer to buy or sell any security. Fund data, distribution history, performance figures and tax characterizations are drawn from Liberty All-Star Equity Fund's Form N-CSR for the year ended December 31, 2025, its Form N-CSR for the year ended December 31, 2021, and issuer press releases dated July 6, 2026 and August 14, 2026. NAV, market price and discount figures are as of July 31, 2026 and are stale by the time you read this. Annualized return calculations on the hypothetical $10,000 illustration, the $0.68 distribution run rate, the payout-on-market-price figure and the reinvestment dilution conclusion are the author's derivations from fund disclosures, not fund-reported metrics. Distribution rates are not yields, are not guaranteed, and the Fund has reduced its distribution rate before. Estimated sources of distribution are not final tax character, which is determined on Form 1099-DIV after year end. Closed-end funds frequently trade at discounts to net asset value and those discounts can widen indefinitely. Past performance does not guarantee future results. Verify all figures against primary filings before acting and consult a licensed financial professional regarding your specific situation.